Medulloblastoma (MB) is the most common malignant brain tumor in children. Patients whose tumors exhibit large cell/anaplastic (LCA) histology usually fail therapy and die from their disease. Improved approaches to treating these patients are likely to come from a deeper understanding of LCA tumors, including their aggressive growth properties and their invasive and metastatic behavior. Unfortunately, human LCA MB tissue is difficult to obtain, and existing genetically engineered mouse (GEM) models of MB rarely display anaplasia or metastasis. To address this problem, we have collected >100 human LCA MBs, including paired samples of primary/metastatic tumors. In addition, we have generated GEM and transplant-based models of LCA MB, and mobilized the transposable element Sleeping Beauty (SB) to promote anaplasia and metastasis in these models. Through analysis of our human and murine datasets, we propose to identify the cells and genes that drive progression in LCA medulloblastoma. This application brings together investigators from three institutions, with collective experience in neural development, stem cell biology and genomics of both human and murine MB. Our complementary backgrounds and expertise uniquely position us to investigate the cellular and molecular basis of LCA MB, and will ultimately allow us to develop more effective approaches to targeting these aggressive tumors. PUBLIC HEALTH RELEVANCE: The proposed studies focus on large cell/anaplastic medulloblastoma, a highly malignant pediatric brain tumor with an extremely poor prognosis. By identifying the cells and genes responsible for the aggressive behavior of this tumor, we hope to develop more effective approaches to therapy.